Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7fb3184003c8>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7fb318326f98>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.0.0
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    # TODO: Implement Function
    real_dim = (image_width, image_height, image_channels)
    inputs_real = tf.placeholder(tf.float32, (None, *real_dim), name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')
    learning_rate = tf.placeholder(tf.float32, None, name='learning_rate')

    return inputs_real, inputs_z, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variabes in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the generator, tensor logits of the generator).

def LeakyReLU(x, alpha=0.2):
    return tf.maximum(alpha * x, x)

def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param image: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    # TODO: Implement Function
    # Using dropouts in discrimnator so as to weaken it's learning model for the data distribution 
    # as well as help generalise better.
    keep_probability = 0.8
    
    with tf.variable_scope('discriminator', reuse=reuse):
        # Input layer is 28x28x1 for MNIST or 28x28x3 for CelebA
        # This is the 1st layer
        # No batch normalization on first layer
        x1 = tf.layers.conv2d(images, 128, 5, strides=2, padding='same', 
                              kernel_initializer=tf.contrib.layers.xavier_initializer())
        x1 = tf.nn.dropout(x1, keep_probability)
        relu1 = LeakyReLU(x1)
        # 14x14x128 now
        #print('discriminator x1 shape ', x1.shape)
        
        # This is the 2nd layer
        x2 = tf.layers.conv2d(relu1, 256, 5, strides=2, padding='same', 
                              kernel_initializer=tf.contrib.layers.xavier_initializer())
        x2 = tf.nn.dropout(x2, keep_probability)
        bn2 = tf.layers.batch_normalization(x2, training=True)
        relu2 = LeakyReLU(bn2)
        # 7x7x256 now
        #print('discriminator x2 shape ', x2.shape)
        
        # This is the 3rd layer
        x3 = tf.layers.conv2d(relu2, 512, 5, strides=2, padding='same', 
                              kernel_initializer=tf.contrib.layers.xavier_initializer())
        x3 = tf.nn.dropout(x3, keep_probability)
        bn3 = tf.layers.batch_normalization(x3, training=True)
        relu3 = LeakyReLU(bn3)
        # 4x4x512 now
        #print('discriminator x3 shape ', x3.shape)

        # This is the 4th fully-connected layer
        # Flatten it
        #flat = tf.reshape(relu2, (-1, 7*7*128)) before adding additional layer
        flat = tf.reshape(relu3, (-1, 4*4*512))
        logits = tf.layers.dense(flat, 1, kernel_initializer=tf.truncated_normal_initializer(stddev=0.02))
        logits = tf.nn.dropout(logits, keep_probability)
        out = tf.sigmoid(logits)
    
    return out, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variabes in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    # TODO: Implement Function
    keep_probability = 0.9 # eliminating dropout weakens generalization in Generator
    # print(out_channel_dim) <- value of 5 in unit test
    #output_dim = (28, 28, out_channel_dim)
    
    with tf.variable_scope('generator', reuse=not is_train): # <- not sure if reuse during training
        # 1st fully connected layer
        x1 = tf.layers.dense(z, 7*7*512, kernel_initializer=tf.truncated_normal_initializer(stddev=0.02))
        # Reshape it to start the convolutional stack
        x1 = tf.reshape(x1, (-1, 7, 7, 256))
        #x1 = tf.nn.dropout(x1, keep_probability)
        x1 = tf.layers.batch_normalization(x1, training=is_train)
        x1 = LeakyReLU(x1)
        # 7x7x512 now
        
        # This is the 2nd layer
        x2 = tf.layers.conv2d_transpose(x1, 256, 5, strides=2, padding='same', 
                                        kernel_initializer=tf.contrib.layers.xavier_initializer())
        #x2 = tf.nn.dropout(x2, keep_probability)
        x2 = tf.layers.batch_normalization(x2, training=is_train)
        x2 = LeakyReLU(x2)
        # 14x14x256 now
        
        # This is the 3rd layer
        x3 = tf.layers.conv2d_transpose(x2, 128, 5, strides=2, padding='same', 
                                        kernel_initializer=tf.contrib.layers.xavier_initializer())
        #x3 = tf.nn.dropout(x3, keep_probability)
        x3 = tf.layers.batch_normalization(x3, training=is_train)
        x3 = LeakyReLU(x3)
        # 28x28x128 now
        
        # This is the 4th, output layer
        # stides=1 because no upscaling
        logits = tf.layers.conv2d_transpose(x3, out_channel_dim, 5, strides=1, padding='same', 
                                            kernel_initializer=tf.contrib.layers.xavier_initializer())
        #logits = tf.nn.dropout(logits, keep_probability)
        # 28x28xout_channel_dim now        
        out = tf.tanh(logits)
    
    return out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    # TODO: Implement Function
    g_model = generator(input_z, out_channel_dim)
    d_model_real, d_logits_real = discriminator(input_real)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)
    
    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real)))
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))
    
    d_loss = d_loss_real + d_loss_fake
    
    return d_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # TODO: Implement Function
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)
    
    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # TODO: Build Model
    #print(data_shape) # <- (6000, 28, 28, 1) for MNIST training
    #print(data_image_mode) # <- "L" for MNIST training
    
    if (data_image_mode == "L"):
        out_channel_dim=1
    else:
        out_channel_dim=3
    
    #tf.reset_default_graph()
    # model_inputs(image_width, image_height, image_channels, z_dim)
    input_real, input_z, lr = model_inputs(data_shape[1], data_shape[2], data_shape[3], z_dim)
    lr = learning_rate
    d_loss, g_loss = model_loss(input_real, input_z, out_channel_dim)
    d_opt, g_opt = model_opt(d_loss, g_loss, lr, beta1)
    
    print_every=10
    show_every=100
    samples, losses = [], []
    steps=0
    
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                # TODO: Train Model
                steps += 1
                
                # print(batch_images.shape) # <- (128, 28, 28, 1) for MNIST
                #batch_images = batch_images.reshape(batch_size, data_shape[1], data_shape[2], data_shape[3])
                # The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5
                #  of 28x28 dimensional images.
                batch_images = batch_images * 2
                
                # Sample random noise for G
                batch_z=np.random.uniform(-1, 1, size=(batch_size, z_dim))
                
                # Run optimizers
                _ = sess.run(d_opt, feed_dict={input_real: batch_images, input_z: batch_z})
                _ = sess.run(g_opt, feed_dict={input_z: batch_z, input_real: batch_images})
                #_ = sess.run(g_opt, feed_dict={input_z: batch_z, input_real: batch_images}) saw suggestion to run 2x
                
                if steps % print_every == 0:
                    # At the end of each epoch, get the losses and print them out
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z})

                    print("Epoch {}/{}...".format(epoch_i+1, epochs),
                          "Discriminator Loss: {:.4f} ...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))
                    # Save losses to view after training
                    losses.append((train_loss_d, train_loss_g))

                if steps % show_every == 0:
                    #show_generator_output(sess, 1, input_z, data_shape[3], data_image_mode)
                    show_generator_output(sess, 9, input_z, data_shape[3], data_image_mode)
                    
        # show images at end of training
        #show_generator_output(sess, 1, input_z, data_shape[3], data_image_mode)
        show_generator_output(sess, 9, input_z, data_shape[3], data_image_mode)

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 64
z_dim = 100
learning_rate = 0.002
beta1 = 0.1

# https://discussions.udacity.com/t/project-facegeneration-indexerror-list-index-out-of-range/246401
tf.reset_default_graph()

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

Epoch 1/2... Discriminator Loss: 0.2654 ... Generator Loss: 32.2275
Epoch 1/2... Discriminator Loss: 0.3249 ... Generator Loss: 26.9800
Epoch 1/2... Discriminator Loss: 0.2816 ... Generator Loss: 29.3915
Epoch 1/2... Discriminator Loss: 5.3296 ... Generator Loss: 6.7937
Epoch 1/2... Discriminator Loss: 2.1493 ... Generator Loss: 6.5792
Epoch 1/2... Discriminator Loss: 9.1840 ... Generator Loss: 25.9268
Epoch 1/2... Discriminator Loss: 0.6548 ... Generator Loss: 1.7897
Epoch 1/2... Discriminator Loss: 1.2082 ... Generator Loss: 8.7612
Epoch 1/2... Discriminator Loss: 0.5177 ... Generator Loss: 3.7103
Epoch 1/2... Discriminator Loss: 1.4380 ... Generator Loss: 3.6875

Epoch 1/2... Discriminator Loss: 1.3329 ... Generator Loss: 2.5693
Epoch 1/2... Discriminator Loss: 2.5640 ... Generator Loss: 5.0681
Epoch 1/2... Discriminator Loss: 2.5882 ... Generator Loss: 2.6396
Epoch 1/2... Discriminator Loss: 1.5877 ... Generator Loss: 1.5120
Epoch 1/2... Discriminator Loss: 1.9613 ... Generator Loss: 1.4544
Epoch 1/2... Discriminator Loss: 1.7848 ... Generator Loss: 0.5033
Epoch 1/2... Discriminator Loss: 1.5491 ... Generator Loss: 0.5930
Epoch 1/2... Discriminator Loss: 1.4484 ... Generator Loss: 1.2203
Epoch 1/2... Discriminator Loss: 0.9405 ... Generator Loss: 1.0150
Epoch 1/2... Discriminator Loss: 0.9409 ... Generator Loss: 2.5434

Epoch 1/2... Discriminator Loss: 1.0912 ... Generator Loss: 1.1021
Epoch 1/2... Discriminator Loss: 1.6363 ... Generator Loss: 0.4267
Epoch 1/2... Discriminator Loss: 2.3692 ... Generator Loss: 0.2945
Epoch 1/2... Discriminator Loss: 2.1856 ... Generator Loss: 0.2709
Epoch 1/2... Discriminator Loss: 1.1055 ... Generator Loss: 1.3071
Epoch 1/2... Discriminator Loss: 1.4402 ... Generator Loss: 1.2880
Epoch 1/2... Discriminator Loss: 1.6654 ... Generator Loss: 0.4360
Epoch 1/2... Discriminator Loss: 1.4509 ... Generator Loss: 0.4542
Epoch 1/2... Discriminator Loss: 1.5587 ... Generator Loss: 0.5217
Epoch 1/2... Discriminator Loss: 1.6968 ... Generator Loss: 0.4342

Epoch 1/2... Discriminator Loss: 1.5483 ... Generator Loss: 0.4395
Epoch 1/2... Discriminator Loss: 1.3796 ... Generator Loss: 0.6631
Epoch 1/2... Discriminator Loss: 1.7698 ... Generator Loss: 0.4508
Epoch 1/2... Discriminator Loss: 2.1197 ... Generator Loss: 0.3147
Epoch 1/2... Discriminator Loss: 2.3250 ... Generator Loss: 0.2358
Epoch 1/2... Discriminator Loss: 1.4718 ... Generator Loss: 0.4611
Epoch 1/2... Discriminator Loss: 0.6841 ... Generator Loss: 1.2887
Epoch 1/2... Discriminator Loss: 1.2306 ... Generator Loss: 2.2898
Epoch 1/2... Discriminator Loss: 1.1506 ... Generator Loss: 1.8497
Epoch 1/2... Discriminator Loss: 0.8751 ... Generator Loss: 0.9558

Epoch 1/2... Discriminator Loss: 1.5784 ... Generator Loss: 1.3364
Epoch 1/2... Discriminator Loss: 1.4978 ... Generator Loss: 1.1486
Epoch 1/2... Discriminator Loss: 1.4928 ... Generator Loss: 1.6595
Epoch 1/2... Discriminator Loss: 1.1615 ... Generator Loss: 2.8096
Epoch 1/2... Discriminator Loss: 0.7983 ... Generator Loss: 1.0567
Epoch 1/2... Discriminator Loss: 1.1006 ... Generator Loss: 0.7524
Epoch 1/2... Discriminator Loss: 2.4836 ... Generator Loss: 0.2606
Epoch 1/2... Discriminator Loss: 1.4212 ... Generator Loss: 0.9790
Epoch 1/2... Discriminator Loss: 1.6233 ... Generator Loss: 2.1987
Epoch 1/2... Discriminator Loss: 1.4925 ... Generator Loss: 0.5120

Epoch 1/2... Discriminator Loss: 1.1186 ... Generator Loss: 0.8559
Epoch 1/2... Discriminator Loss: 1.4115 ... Generator Loss: 1.0557
Epoch 1/2... Discriminator Loss: 1.4767 ... Generator Loss: 1.4023
Epoch 1/2... Discriminator Loss: 1.1456 ... Generator Loss: 1.2576
Epoch 1/2... Discriminator Loss: 1.9574 ... Generator Loss: 1.3772
Epoch 1/2... Discriminator Loss: 1.6291 ... Generator Loss: 0.4809
Epoch 1/2... Discriminator Loss: 1.4197 ... Generator Loss: 0.8903
Epoch 1/2... Discriminator Loss: 1.3386 ... Generator Loss: 1.3950
Epoch 1/2... Discriminator Loss: 1.2415 ... Generator Loss: 1.0511
Epoch 1/2... Discriminator Loss: 1.3935 ... Generator Loss: 0.4840

Epoch 1/2... Discriminator Loss: 1.4066 ... Generator Loss: 0.9027
Epoch 1/2... Discriminator Loss: 1.2537 ... Generator Loss: 0.9032
Epoch 1/2... Discriminator Loss: 1.6710 ... Generator Loss: 1.5763
Epoch 1/2... Discriminator Loss: 1.2922 ... Generator Loss: 0.8645
Epoch 1/2... Discriminator Loss: 1.6379 ... Generator Loss: 1.7832
Epoch 1/2... Discriminator Loss: 1.5040 ... Generator Loss: 1.6985
Epoch 1/2... Discriminator Loss: 1.2859 ... Generator Loss: 0.6740
Epoch 1/2... Discriminator Loss: 1.3283 ... Generator Loss: 0.9822
Epoch 1/2... Discriminator Loss: 1.2895 ... Generator Loss: 1.0500
Epoch 1/2... Discriminator Loss: 1.4276 ... Generator Loss: 1.4086

Epoch 1/2... Discriminator Loss: 1.2529 ... Generator Loss: 0.9991
Epoch 1/2... Discriminator Loss: 1.2888 ... Generator Loss: 1.0349
Epoch 1/2... Discriminator Loss: 1.3230 ... Generator Loss: 0.8259
Epoch 1/2... Discriminator Loss: 1.4283 ... Generator Loss: 1.1442
Epoch 1/2... Discriminator Loss: 1.3894 ... Generator Loss: 0.9153
Epoch 1/2... Discriminator Loss: 1.3507 ... Generator Loss: 0.7992
Epoch 1/2... Discriminator Loss: 1.4154 ... Generator Loss: 0.6804
Epoch 1/2... Discriminator Loss: 1.2047 ... Generator Loss: 0.8473
Epoch 1/2... Discriminator Loss: 1.5258 ... Generator Loss: 1.2014
Epoch 1/2... Discriminator Loss: 1.3035 ... Generator Loss: 1.4701

Epoch 1/2... Discriminator Loss: 1.3496 ... Generator Loss: 0.9898
Epoch 1/2... Discriminator Loss: 1.2579 ... Generator Loss: 0.8580
Epoch 1/2... Discriminator Loss: 1.4216 ... Generator Loss: 1.2486
Epoch 1/2... Discriminator Loss: 1.3097 ... Generator Loss: 0.5803
Epoch 1/2... Discriminator Loss: 1.2845 ... Generator Loss: 1.0545
Epoch 1/2... Discriminator Loss: 1.6251 ... Generator Loss: 1.5960
Epoch 1/2... Discriminator Loss: 1.5169 ... Generator Loss: 0.5304
Epoch 1/2... Discriminator Loss: 1.7124 ... Generator Loss: 0.4419
Epoch 1/2... Discriminator Loss: 1.3658 ... Generator Loss: 0.7402
Epoch 1/2... Discriminator Loss: 1.3068 ... Generator Loss: 0.7198

Epoch 1/2... Discriminator Loss: 1.3892 ... Generator Loss: 1.0280
Epoch 1/2... Discriminator Loss: 1.4644 ... Generator Loss: 0.4730
Epoch 1/2... Discriminator Loss: 1.3640 ... Generator Loss: 1.0644
Epoch 2/2... Discriminator Loss: 1.3053 ... Generator Loss: 0.6800
Epoch 2/2... Discriminator Loss: 1.3524 ... Generator Loss: 0.5528
Epoch 2/2... Discriminator Loss: 1.3248 ... Generator Loss: 0.7206
Epoch 2/2... Discriminator Loss: 1.5889 ... Generator Loss: 1.0679
Epoch 2/2... Discriminator Loss: 1.3587 ... Generator Loss: 0.9260
Epoch 2/2... Discriminator Loss: 1.3622 ... Generator Loss: 0.9587
Epoch 2/2... Discriminator Loss: 1.6212 ... Generator Loss: 0.4174

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 64
z_dim = 100
learning_rate = 0.0002
beta1 = 0.1


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 7

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Epoch 1/7... Discriminator Loss: 0.4713 ... Generator Loss: 3.9013
Epoch 1/7... Discriminator Loss: 0.5393 ... Generator Loss: 4.0696
Epoch 1/7... Discriminator Loss: 0.5885 ... Generator Loss: 3.2914
Epoch 1/7... Discriminator Loss: 1.1294 ... Generator Loss: 3.9136
Epoch 1/7... Discriminator Loss: 0.9016 ... Generator Loss: 3.8389
Epoch 1/7... Discriminator Loss: 0.8217 ... Generator Loss: 1.1937
Epoch 1/7... Discriminator Loss: 0.9056 ... Generator Loss: 1.4941
Epoch 1/7... Discriminator Loss: 0.7855 ... Generator Loss: 1.5963
Epoch 1/7... Discriminator Loss: 2.9813 ... Generator Loss: 0.2236
Epoch 1/7... Discriminator Loss: 0.6984 ... Generator Loss: 4.9919

Epoch 1/7... Discriminator Loss: 0.4409 ... Generator Loss: 2.6330
Epoch 1/7... Discriminator Loss: 0.9357 ... Generator Loss: 5.2556
Epoch 1/7... Discriminator Loss: 1.6038 ... Generator Loss: 4.9961
Epoch 1/7... Discriminator Loss: 3.7294 ... Generator Loss: 6.6898
Epoch 1/7... Discriminator Loss: 0.6209 ... Generator Loss: 1.4794
Epoch 1/7... Discriminator Loss: 2.0663 ... Generator Loss: 0.3924
Epoch 1/7... Discriminator Loss: 1.1817 ... Generator Loss: 0.6790
Epoch 1/7... Discriminator Loss: 0.8945 ... Generator Loss: 0.8799
Epoch 1/7... Discriminator Loss: 0.7201 ... Generator Loss: 1.3308
Epoch 1/7... Discriminator Loss: 1.6070 ... Generator Loss: 0.3786

Epoch 1/7... Discriminator Loss: 1.6508 ... Generator Loss: 0.4648
Epoch 1/7... Discriminator Loss: 2.2274 ... Generator Loss: 0.2588
Epoch 1/7... Discriminator Loss: 1.4058 ... Generator Loss: 0.5651
Epoch 1/7... Discriminator Loss: 1.3325 ... Generator Loss: 0.5434
Epoch 1/7... Discriminator Loss: 1.9060 ... Generator Loss: 0.3291
Epoch 1/7... Discriminator Loss: 1.5142 ... Generator Loss: 5.1636
Epoch 1/7... Discriminator Loss: 2.2582 ... Generator Loss: 3.5315
Epoch 1/7... Discriminator Loss: 0.8843 ... Generator Loss: 1.9896
Epoch 1/7... Discriminator Loss: 1.0612 ... Generator Loss: 2.7415
Epoch 1/7... Discriminator Loss: 1.1434 ... Generator Loss: 0.5962

Epoch 1/7... Discriminator Loss: 0.8376 ... Generator Loss: 1.0758
Epoch 1/7... Discriminator Loss: 0.4655 ... Generator Loss: 4.1071
Epoch 1/7... Discriminator Loss: 0.6216 ... Generator Loss: 1.4840
Epoch 1/7... Discriminator Loss: 0.4112 ... Generator Loss: 2.2271
Epoch 1/7... Discriminator Loss: 0.9110 ... Generator Loss: 1.0538
Epoch 1/7... Discriminator Loss: 0.6438 ... Generator Loss: 2.6285
Epoch 1/7... Discriminator Loss: 1.3481 ... Generator Loss: 0.7775
Epoch 1/7... Discriminator Loss: 1.0942 ... Generator Loss: 0.6925
Epoch 1/7... Discriminator Loss: 1.6282 ... Generator Loss: 4.7294
Epoch 1/7... Discriminator Loss: 0.9818 ... Generator Loss: 1.1558

Epoch 1/7... Discriminator Loss: 0.7890 ... Generator Loss: 1.1326
Epoch 1/7... Discriminator Loss: 1.3001 ... Generator Loss: 0.6169
Epoch 1/7... Discriminator Loss: 0.6504 ... Generator Loss: 1.3371
Epoch 1/7... Discriminator Loss: 0.5839 ... Generator Loss: 1.4342
Epoch 1/7... Discriminator Loss: 1.0082 ... Generator Loss: 1.7498
Epoch 1/7... Discriminator Loss: 2.1116 ... Generator Loss: 2.9668
Epoch 1/7... Discriminator Loss: 1.2447 ... Generator Loss: 0.6714
Epoch 1/7... Discriminator Loss: 0.6554 ... Generator Loss: 1.8945
Epoch 1/7... Discriminator Loss: 0.4713 ... Generator Loss: 3.5441
Epoch 1/7... Discriminator Loss: 1.2293 ... Generator Loss: 0.5851

Epoch 1/7... Discriminator Loss: 1.8688 ... Generator Loss: 0.4120
Epoch 1/7... Discriminator Loss: 0.6167 ... Generator Loss: 2.8166
Epoch 1/7... Discriminator Loss: 0.7463 ... Generator Loss: 3.8027
Epoch 1/7... Discriminator Loss: 1.1147 ... Generator Loss: 1.6056
Epoch 1/7... Discriminator Loss: 0.9710 ... Generator Loss: 2.7370
Epoch 1/7... Discriminator Loss: 1.1211 ... Generator Loss: 0.7025
Epoch 1/7... Discriminator Loss: 1.0313 ... Generator Loss: 0.8168
Epoch 1/7... Discriminator Loss: 0.6679 ... Generator Loss: 1.4238
Epoch 1/7... Discriminator Loss: 0.5026 ... Generator Loss: 2.8807
Epoch 1/7... Discriminator Loss: 0.9945 ... Generator Loss: 0.7373

Epoch 1/7... Discriminator Loss: 0.8261 ... Generator Loss: 1.0580
Epoch 1/7... Discriminator Loss: 0.9606 ... Generator Loss: 0.8296
Epoch 1/7... Discriminator Loss: 0.6238 ... Generator Loss: 1.2102
Epoch 1/7... Discriminator Loss: 1.9954 ... Generator Loss: 0.2635
Epoch 1/7... Discriminator Loss: 1.1260 ... Generator Loss: 0.7617
Epoch 1/7... Discriminator Loss: 0.9508 ... Generator Loss: 0.8455
Epoch 1/7... Discriminator Loss: 1.2919 ... Generator Loss: 0.4965
Epoch 1/7... Discriminator Loss: 1.0039 ... Generator Loss: 1.1031
Epoch 1/7... Discriminator Loss: 1.1475 ... Generator Loss: 0.6389
Epoch 1/7... Discriminator Loss: 0.5013 ... Generator Loss: 1.7711

Epoch 1/7... Discriminator Loss: 1.2488 ... Generator Loss: 2.0340
Epoch 1/7... Discriminator Loss: 0.8642 ... Generator Loss: 4.0315
Epoch 1/7... Discriminator Loss: 1.1876 ... Generator Loss: 0.5735
Epoch 1/7... Discriminator Loss: 1.3841 ... Generator Loss: 1.8713
Epoch 1/7... Discriminator Loss: 0.5575 ... Generator Loss: 1.9560
Epoch 1/7... Discriminator Loss: 0.7431 ... Generator Loss: 1.6246
Epoch 1/7... Discriminator Loss: 1.0602 ... Generator Loss: 0.7740
Epoch 1/7... Discriminator Loss: 1.4195 ... Generator Loss: 0.5007
Epoch 1/7... Discriminator Loss: 1.2572 ... Generator Loss: 0.6418
Epoch 1/7... Discriminator Loss: 0.4804 ... Generator Loss: 2.5659

Epoch 1/7... Discriminator Loss: 0.7702 ... Generator Loss: 2.4004
Epoch 1/7... Discriminator Loss: 1.5316 ... Generator Loss: 2.6433
Epoch 1/7... Discriminator Loss: 0.9009 ... Generator Loss: 0.8225
Epoch 1/7... Discriminator Loss: 0.5895 ... Generator Loss: 1.6727
Epoch 1/7... Discriminator Loss: 1.6999 ... Generator Loss: 0.3394
Epoch 1/7... Discriminator Loss: 1.3212 ... Generator Loss: 1.6049
Epoch 1/7... Discriminator Loss: 0.6368 ... Generator Loss: 2.0648
Epoch 1/7... Discriminator Loss: 0.5940 ... Generator Loss: 1.5543
Epoch 1/7... Discriminator Loss: 1.9397 ... Generator Loss: 0.3201
Epoch 1/7... Discriminator Loss: 0.8092 ... Generator Loss: 1.2448

Epoch 1/7... Discriminator Loss: 1.5135 ... Generator Loss: 2.4910
Epoch 1/7... Discriminator Loss: 1.8981 ... Generator Loss: 0.2961
Epoch 1/7... Discriminator Loss: 0.6094 ... Generator Loss: 1.5997
Epoch 1/7... Discriminator Loss: 1.1552 ... Generator Loss: 2.5899
Epoch 1/7... Discriminator Loss: 0.7654 ... Generator Loss: 2.9577
Epoch 1/7... Discriminator Loss: 0.7325 ... Generator Loss: 2.6202
Epoch 1/7... Discriminator Loss: 1.1471 ... Generator Loss: 0.7816
Epoch 1/7... Discriminator Loss: 1.2702 ... Generator Loss: 0.7774
Epoch 1/7... Discriminator Loss: 0.7768 ... Generator Loss: 2.3073
Epoch 1/7... Discriminator Loss: 0.8769 ... Generator Loss: 0.9705

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.